- Email: [email protected]
Erythrocyte ouabain binding and intracellular Na+ in normotensive obese women and obese women receiving medication for hypertension
BIOCHEMICAL
32, 232-241
MEDICINE
t 1984)
Erythrocyte Ouabain Binding and Intracellular Na’ in Normotensive Obese Women and Obese Women Receiving Medication for Hypertension DOUGLAS
P.
WEBSTER,
LON
J.
VAN
WINKLE,
AND
JOHN
J.
KARRAT
Chicago College of Osteopathic Medicine, 1122 East 53rd Street, Chicago. Illinois 60615 Received
October
24. 1983
De Luise and collaborators suggested that at least two populations of obese people can be distinguished biochemically (1,2). One group, described as having primary obesity, has low levels of Na’-K+-ATPase associated with their erythrocyte membranes. In contrast, another group which consists of people with obesity secondary to a pathological condition, such as a central nervous system lesion, tends to have normal or slightly elevated numbers of Na’-K’-ATPase sites on their red cells. In other studies, Garay et al. found a diminished capacity for Naf extrusion from erythrocytes, via a Na’-K’ cotransport system, among hypertensive people and their first-degree relatives (3). They propose that this defect, if also present in the smooth muscle cells of arterioles, could lead to increased intracellular Na’, and therefore increased intracellular Ca”, via a ubiquitous Na’-Ca” exchange system. An increase in tonicity of this tissue, due to a higher intracellular Ca” concentration, might then result in hypertension, as proposed by Blaustein (4). In contrast, Heagerty and associates have concluded that a diminished capacity to extrude intracellular Na’ may be a marker for, but does not participate directly in producing, raised blood pressure (5). They argue that if diminished Na’ extrusion leads to hypertension, then first-degree relatives of hypertensive patients should also have high blood pressure when they have this defect (5). However, it is possible that other factors compensate, at first, for diminished Na’ extrusion, and that hypertension is manifest when these compensatory mechanisms break down. We suggested that hypertension may be found more often among obese people because persons with primary obesity may have a diminished capacity to extrude Na’ from their cells (6). The present study was performed to determine if two populations of Na’-K’-ATPase levels 232 0006-2944/84 $3 .OO Copyright 0 1984 by Academic All tights
of reproduction
Press. Inc. in any form reserved
Na’
, OBESITY,
AND
HYPERTENSION
233
exist among a relatively homogeneous population of obese people (i.e., obese, Black women in Chicago). In addition, we wanted to ascertain if an inverse relationship exists between Na+-K+-ATPase levels and blood pressure in these people. Finally, we wanted to determine if the latter relationship is more pronounced among obese patients also receiving medication for hypertension. These people are of interest since their inherent ability to regulate blood pressure (7) appears to be poor, as evidenced by their need for treatment. Thus, they might be less able to compensate for differences in their capacity to extrude Na+ from cells. MATERIALS
AND METHODS
Red blood cells from 55 volunteers entering the Optifast Weight Reduction Program at the Chicago Osteopathic Medical Center were analyzed for [3H]ouabain binding, using a modification of a procedure described previously (1,2). Intracellular Na+ concentrations were determined by atomic absorption spectroscopy. All patients were obese, Black women with normal thyroid and blood profiles. Each gave informed consent for the procedures. The mean (+ standard deviation) age, body mass, and height of the 42 normotensive, obese patients were 32 t 9 years, 91 +- 14 kg, and 153 + 14 cm, respectively. Thirteen subjects (44 2 12 years, 103 2 17 kg, 152 + 12 cm) were receiving hypertensive medication at the time of their participation in the study (eight, methyldopa; two, prazosin; one, methyldopa + prazosin; one, chlorthalidone; one, a triamterene/ hydrochlorothiazide combination). Blood was collected in heparinized Vacutainer tubes at the time of the subjects’ initial visit to our clinic, and where possible, after the patient had dieted for 4 to 6 weeks. Isotonic saline (10 ml) was added to 3 ml of whole blood which was centrifuged and the plasma and saline removed. Following three similar washings with 140 mM choline chloride, the red cells were separated from the final wash by centrifugation at 1700g for 5 min. Two hundred microliters of the resultant “packed” erythrocytes were removed and added to 3.8 ml of buffer (150 mM NaCl, 30 mM Hepes, and 10 mM dextrose; pH = 7.4). Four hundred-microliter aliquots of the latter suspension were placed in each of two small glass tubes. Fifty microliters of 500 nM ouabain in buffer was added to one tube while the same volume of 500 PM ouabain in buffer was added to the other tube. Both of these ouabain solutions also contained 1.0 pmole of 13H]ouabain (17 Ci/mmole, Amersham). The tubes were incubated at 37°C for 60 min with gentle agitation. The cells were then washed three times with ice-cold choline chloride and their bound radioactivity eluted with 300 ~1 of 12% trichloroacetic acid. The difference in counts per minute by red cells exposed to the lower (55 nM) versus the higher (55 FM) concentration of ouabain was taken as a measure of ouabain bound
234
WEBSTER,
VAN
WINKLE.
AND
KARRAT
specifically to Na’-K ‘-ATPase. A counting efficiency of 42% was used to calculate the moles of ouabain bound to the Na’-K’-ATPase contained in 20 ~1 of erythrocytes. Erythrocyte intracellular Na’ concentrations were determined by lysing 100 ~1 of the “packed” red cells prepared above in 9.9 ml of 3 PM lithium nitrate solution, then analyzing the preparation by flame atomic absorption spectroscopy. Blood pressures were measured using a conventional mercury column sphygmomanometer. Correlation coefficients were calculated to detect statistically significant correlations among ouabain binding, intracellular Na’ , and diastolic blood pressure. One-tailed tests were considered appropriate since the expected relationships were formulated before the study was begun. Changes in ouabain binding, intracellular Na’ , and diastolic blood pressure after dieting were analyzed statistically using two-tailed pairing design t tests. We used the x’ goodness of fit test to determine if data were distributed normally (8). RESULTS
The ouabain binding level was inversely correlated with diastolic blood pressure (Fig. 1: r = -0.59) and intracellular Na’ concentration (Fig. 2; r = -0.68) in patients receiving medication for their hypertension. Intracellular Nat correlated directly (r = +0.68) with diastolic blood pressure in the latter patients (Fig. 3). In contrast, there was only a weak inverse correlation (r = - 0.26) between ouabain binding and intracellular Na+ (Fig. 2), and no correlation between diastolic blood pressure and intracellular Na+ or ouabain binding among nonmedicated, normotensive obese patients (Figs. 1 and 3). This lack of relationship among the parameters was not altered when data from the 15 heaviest, 16 oldest, or 11 “heaviest and oldest” normotensive patients were considered in separate analyses. Consistent with the preceding observations, medicated patients had lower diastolic blood pressure (P < 0.05) and intracellular Na’ (P < 0.05) and higher ouabain binding (P < 0.02) after dieting. These parameters were not affected by dieting in normotensive patients (Table I). However, two populations of ouabain binding were found among the latter group (Fig. 4). DISCUSSION Blood pressure and Naf transport in normotensive patients. If erythrocyte ouabain binding reflects a level of ouabain binding characteristic of an individual, then there may be no relationship between blood pressure and smooth muscle cell Na’-K+-ATPase levels among normotensive, obese patients (Fig. 1). If Na’-K’-ATPase levels are to contribute to blood pressure via the mechanism proposed by Blaustein (4), they must
Na’ . OBESITY,
235
AND HYPERTENSION
0
60
n 10
20
n.
30 Ouobain
40 binding
(tmO~/2Oul
50
60
cells1
FIG. 1. Relationship between ouabain binding and diastolic blood pressure in obese women. Women receiving medication for hypertension (0; r = -0.59; P < 0.02; onetailed test); normotensive patients (0; r = +0.12). One (0) point (48 mm Hg) is not shown. The fine is for medicated patients.
Sect the intracellular Na+ concentration. However, only a weak inverse correlation was observed between ouabain binding and intracellular Na’ concentration (Fig. 2), suggesting that other Na+ transport mechanisms may partially compensate for the potential effect of Na’-K’-ATPase levels on intracellular Na+. Moreover, since the intracellular Na+ con-
6 90
I 40
Al 3a Ouabain
biding
n
(fmd
I 50
. w
/20uicelb)
FIG. 2. Relationship between ouabain binding and intracellular Na’ in obese women. Women receiving medication for hypertension (A; r = -0.68; P < 0.01); normotensive patients (0; r = -0.26; P < 0.05; one-tailed test). The line is for medicated patients.
236
WEBSTER,
VAN WINKLE,
lnlracellular
AND KARRAT
No+ (mmol / I cells)
FIG. 3. Relationship between intracellular Na- and diastolic blood pressure in obese women. Women receiving medication for hypertension (+; r = +0.68; P < 0.01); normotensive patients (0; r = -0.01). One (0) point (48 mm Hg) is not shown. The line is for medicated patients.
centration is unrelated to diastolic blood pressure among normotensive patients (Fig. 3), other factors may compensate for the residual effect of Na’-K+-ATPase levels on intracellular Na+, thus completely compensating for the potential effect of the enzyme levels on blood pressure (Fig. 1). Blood pressure and Na’ transport in hypertesive patients receiving treatment. A strong inverse correlation between ouabain binding levels
Ouobain
Binding
(fmol / 20~1 cells)
FIG. 4. Distribution of erythrocyte ouabain binding among obese women. The levels of ouabain binding were not normally distributed among normotensive women (open bars; P < 0.05). Cross hatching indicates patients receiving hypertensive medication.
Na’,
OBESITY,
AND
HYPERTENSION
237
and intracellular Na+ concentration was seen in obese patients receiving medication for hypertension (Fig. 2). Intracellular Na+ levels were strongly correlated with blood pressure (Fig. 3) and, as expected from these two relationships, ouabain binding was inversely correlated with blood pressure (Fig. 1). Thus it appears that obese patients receiving medication for hypertension may be less able to compensate (a) for an effect of Na+K+-ATPase levels on intracellular Na’ and (b) for an effect of intracellular Na’ levels on blood pressure. Consistent with this hypothesis, an increase in ouabain binding was accompanied by a decrease in both intracellular Na’ and blood pressure when medicated patients dieted (Table 1). One might theorize that dieting reduces Na’ intake and that lower Na’ consumption decreases production of an inhibitor of Na’-K+-ATPase (9); thus affecting ouabain binding levels in some way (other explanations for this increase in ouabain binding are also possible, as discussed below). Regardless of the reason(s) for the changes, however, an increase in ouabain binding was associated with a decrease in intracellular Na+ and blood pressure in each of the medicated patients. In contrast, only half of the normotensive patients from whom blood samples were available after dieting had changes in blood pressure and intracellular Na’ in the opposite direction of the change in their level of erythrocyte ouabain binding. This occurred in spite of the fact that the changes in ouabain binding and blood pressure after dieting were of similar absolute magnitude in both medicated and nonmedicated, normotensive patients. Thus, Na’K’-ATPase levels appear to be inversely correlated with intracellular Na’ and blood pressure only in those obese patients also receiving medication for hypertension; possibly because their compensatory mechanisms have broken down. “Primary obesity” as a possible risk factor for hypertension. Our data (Fig. 4) support the hypothesis of De Luise et al. (1,2) that there are at least two populations of Na’-K’-ATPase levels among obese people. One group appears to have a family history of obesity (primary obesity), while another group may be obese as a result of other conditions such as central nervous system lesions (secondary obesity) (1,2). Subjects with primary obesity appear to have lower than normal Na+-K’-ATPase levels which may predispose them not only to obesity (1,2), but to hypertension as well (6). Mir et al. (10) have suggested that variation in NaC-K’-ATPase activity may result, in part, from differences in food intake near the time when blood samples are obtained from subjects. Thus, we may have found two populations of ouabain binding levels among obese people because some patients had eaten before coming to our clinic and some had not. Mir and associates have also presented data that seem to contradict the findings of De Luise et al. (1,2) since they failed to detect diminished
INTRACELLULAR
4.8 2 0.4 4.5 2 0.3
Mean (t SE) duration of dieting (weeks)
BINDING,
TABLE
1
, .AND BLOOD PRESSURF METHYLWPA
0.2 -c 3.7 6.1 k 1.7”
Ouabain binding (fmole/20 ~1 of cells)
NA-
OBLSL
WOMEN
1s OBESF
0.3 2 1.3 -4.6 t 1.7
Intracellular Na* (mM)
-0.7 -7.0
2 5.5 ‘- 2.5’
Wows
-2.7 2 0.8” - 2.4 i 0.7”
Body mass (kg)
RECEIVING
Na-. dias:olic blood binding after dieting
3 years, respectively. 2 year<. respectively.
Diastolic blood pressure (mm Hg)
Mean change after dieting 2 SE
IN NORMOTENSIVE
” These normotensive patients had a mean initial body mass. weight, and age 2 SE of 89 ? 5 kg, 162 +_ 2 cm. and 33 -f * These hypertensive patients had a mean initial body mass, height, and age -+ SE of 100 + S kg. 164 I 2 cm. and 42 -+ P < 0.05. ‘! P c: 0.02 (pairing design f test; 8). Statistical tests were applied to compare parameters (ouabain binding. intracellular pressure. and body mass) before vs after dieting [e.g.. patients receiving methyldopa had significantly higher levels of ouabain (P < 0.02) while this parameter was not affected when nonmedicated patients dieted].
6 6
None” Methyldopa*
ON OUABAIN
Number of patients
OF DIETING
Hypertensive medication
EFFECT
Na’,
OBESITY,
AND
HYPERTENSION
239
Na’-K’-ATPase activity in the erythrocytes of a particular group of obese people, or among obese people in general (11). However, Mir et al. (IO,1 I) studied actual Na’-K+-ATPase activity (i.e., the rate of generation of inorganic phosphate from ATP) while De Luise and collaborators (1,2) studied ouabain binding. The latter is a measure of the number of Na’-K’-ATPase sites, and not an estimation of the ability of these sites to hydrolyze ATP. While it is possible that the capacity to hydrolyze ATP might be subject to modulation following a meal [it increases according to Mir et al., (IO)], it seems unlikely that the number of Na’-K’-ATPase molecules in the red cell membrane would change after eating. As expected, we found no increase in erythrocyte ouabain binding I hr after five, healthy, nonobese subjects each ate a breakfast consisting of two fried eggs; bacon, sausage, or ham: two slices of buttered toast and jelly; and eight ounces of orange juice (data not shown). These observations may explain why we (Fig. 4) and De Luise and associates (1,2) can distinguish two populations of Na+-K’-ATPase levels among obese people while Mir and collaborators ( lo,11 ) are unable to detect two similar populations. Taken together, the data suggest that primary obesity (1,2) is associated with a decrease in Na’-K’-ATPase sites per unit area of membrane. According to our premise, hypertension is more likely to occur in these patients, than in people with higher levels of Na’-Kf-ATPase, if they lose their ability to compensate for the potential effect of Na’-K’ATPase levels on intracellular Na’ and blood pressure. If obese patients receiving medication for hypertension have primary obesity (this is not a requirement of the present hypothesis), then it must be proposed that their medication has increased the number of Na’-K’-ATPase sites on the cell membranes of at least some patients. The medicated patients have ouabain binding levels which are spread across the entire range of binding found among normotensive patients (Fig. 4). In this regard, methyldopa binds tightly to erythrocyte membranes (12) and reverses the inhibition of red cell Naf-K’-ATPase by vanadate (13). Thus, methyldopa can influence Na+-K+-ATPase activity in at least one way. Moreover, if methyldopa affects Na+ -K’-ATPase levels, this drug might be even more effective when patients lose weight (e.g., milligrams of medication per kilograms of body mass would increase). Consistent with this possibility, ouabain binding increased when medicated patients dieted (Table I). Although the dose of methyldopa was reduced for one patient after only 1 week of dieting, ouabain binding was found to increase significantly after dieting whether or not this patient was considered in the statistical analysis. While the present and other (1,2) studies permit formulation of interesting hypotheses, the mechanism by which diminished ouabain binding sites
240
WEBSTER,
VAN WINKLE.
AND KARRAT
on membranes contribute to the development of obesity and/or hypertension must be elucidated before a causal relationship between these parameters can be accepted. This would include not only verification of Blaustein’s proposal (4), but a more complete description of (a) how we compensate for changes in our blood pressure and (b) the consequences when these compensatory mechanisms fail. SUMMARY People with “primary obesity” may be hypertensive because they have lost their ability to compensate for the effect of low Na’-K’-ATPase levels on blood pressure. In obese patients receiving hypertensive medication (n = 13), but not in normotensive nonmedicated patients (n = 42), diastolic blood pressure was inversely correlated with erythrocyte ouabain binding (P < 0.02) and directly correIated with intracellular Na’ concentration (P < 0.01). Moreover, there was a stronger inverse relationship between ouabain binding and intracellular Nat in patients receiving medication for hypertension (P < 0.01) than in normotensive patients (P < 0.05). These data suggest that patients receiving hypertensive medication may be less able to compensate than normotensive patients, (a) for the potential effect of Na’- K’-ATPase levels on intracellular Na+ and (b) for the potential effect of intracellular Na+ concentration on diastolic blood pressure. We propose that obese people with low levels of ouabain binding (primary obesity) may have an increased risk of developing hypertension if their compensatory mechanisms fail. ACKNOWLEDGMENTS We are indebted to Dr. Conrad Naleway of the American Dental Association Research Foundation for use of the atomic absorption spectrophotometer and Mr. Hwai-Nan Chou for technical assistance. We thank Dr. Gary Oltmans for valuable discussions. Thanks to Mr. Allan Campione, Dr. Wells Farnsworth, Barbara Le Breton, and Dr. Daniel Richardson for their help in preparing the manuscript. Supported by Chicago College of Osteopathic Medicine.
REFERENCES 1. De Luise, M., Blackbum, G. L., and Flier, J. S.. N. Engl. J. Med. 303, 1017 (1980). 2. De Luise, M., Rappaport, E., and Flier, J. S., Merabolism 31, 1153 (1982). 3. Garay, R. P., Dagher, G., Pemollet. M.-G,, Devynck, M.-A., and Meyer, P.. Narure (London) 284, 281 (1980). 4. Blaustein, M. P., Amer. J. Physiol. 232, Cl65 (1977). 5. Heagerty, A. M., Bing, R. F., Mimer, M., Thurston, H., and Swales, J. D., Lancer 2, 894 (1982). 6. Van Winkle, L. J., N. Eng/. .I. Med. 304, 358 (1981). 7. Guyton, A. C., ‘*Textbook of Medical Physiology.” Saunders, Philadelphia, 1981. 8. Snedecor, G. W., and Cochran, W. G., ‘*Statistical Methods.” Iowa State Univ. Press, Ames, 1%7.
Na’ , OBESITY,
AND HYPERTENSION
241
9. Marx, J. L., Science (Washington, D.C.) 212, 1255 (1981). 10. Mir, M. A., Charalambous, B. M., and Morgan, K., N. Engl. J. Med. 306, 809 (1982). 11. Mir, M. A., Charalambous, B. M., Morgan, K., and Evans, P. J., N. Engl. J. Med. 305, 1264 (1981). 12. Green, F. A., Jung, C. Y., Rampal, A., and Lorusso. D. J., C/in. Exp. Immunol. 40, 554 (1980). 13. Hudgins, P. M., and Bond, G. H., Fed. Proc. 37, 3 13 (1978).
32, 232-241
MEDICINE
t 1984)
Erythrocyte Ouabain Binding and Intracellular Na’ in Normotensive Obese Women and Obese Women Receiving Medication for Hypertension DOUGLAS
P.
WEBSTER,
LON
J.
VAN
WINKLE,
AND
JOHN
J.
KARRAT
Chicago College of Osteopathic Medicine, 1122 East 53rd Street, Chicago. Illinois 60615 Received
October
24. 1983
De Luise and collaborators suggested that at least two populations of obese people can be distinguished biochemically (1,2). One group, described as having primary obesity, has low levels of Na’-K+-ATPase associated with their erythrocyte membranes. In contrast, another group which consists of people with obesity secondary to a pathological condition, such as a central nervous system lesion, tends to have normal or slightly elevated numbers of Na’-K’-ATPase sites on their red cells. In other studies, Garay et al. found a diminished capacity for Naf extrusion from erythrocytes, via a Na’-K’ cotransport system, among hypertensive people and their first-degree relatives (3). They propose that this defect, if also present in the smooth muscle cells of arterioles, could lead to increased intracellular Na’, and therefore increased intracellular Ca”, via a ubiquitous Na’-Ca” exchange system. An increase in tonicity of this tissue, due to a higher intracellular Ca” concentration, might then result in hypertension, as proposed by Blaustein (4). In contrast, Heagerty and associates have concluded that a diminished capacity to extrude intracellular Na’ may be a marker for, but does not participate directly in producing, raised blood pressure (5). They argue that if diminished Na’ extrusion leads to hypertension, then first-degree relatives of hypertensive patients should also have high blood pressure when they have this defect (5). However, it is possible that other factors compensate, at first, for diminished Na’ extrusion, and that hypertension is manifest when these compensatory mechanisms break down. We suggested that hypertension may be found more often among obese people because persons with primary obesity may have a diminished capacity to extrude Na’ from their cells (6). The present study was performed to determine if two populations of Na’-K’-ATPase levels 232 0006-2944/84 $3 .OO Copyright 0 1984 by Academic All tights
of reproduction
Press. Inc. in any form reserved
Na’
, OBESITY,
AND
HYPERTENSION
233
exist among a relatively homogeneous population of obese people (i.e., obese, Black women in Chicago). In addition, we wanted to ascertain if an inverse relationship exists between Na+-K+-ATPase levels and blood pressure in these people. Finally, we wanted to determine if the latter relationship is more pronounced among obese patients also receiving medication for hypertension. These people are of interest since their inherent ability to regulate blood pressure (7) appears to be poor, as evidenced by their need for treatment. Thus, they might be less able to compensate for differences in their capacity to extrude Na+ from cells. MATERIALS
AND METHODS
Red blood cells from 55 volunteers entering the Optifast Weight Reduction Program at the Chicago Osteopathic Medical Center were analyzed for [3H]ouabain binding, using a modification of a procedure described previously (1,2). Intracellular Na+ concentrations were determined by atomic absorption spectroscopy. All patients were obese, Black women with normal thyroid and blood profiles. Each gave informed consent for the procedures. The mean (+ standard deviation) age, body mass, and height of the 42 normotensive, obese patients were 32 t 9 years, 91 +- 14 kg, and 153 + 14 cm, respectively. Thirteen subjects (44 2 12 years, 103 2 17 kg, 152 + 12 cm) were receiving hypertensive medication at the time of their participation in the study (eight, methyldopa; two, prazosin; one, methyldopa + prazosin; one, chlorthalidone; one, a triamterene/ hydrochlorothiazide combination). Blood was collected in heparinized Vacutainer tubes at the time of the subjects’ initial visit to our clinic, and where possible, after the patient had dieted for 4 to 6 weeks. Isotonic saline (10 ml) was added to 3 ml of whole blood which was centrifuged and the plasma and saline removed. Following three similar washings with 140 mM choline chloride, the red cells were separated from the final wash by centrifugation at 1700g for 5 min. Two hundred microliters of the resultant “packed” erythrocytes were removed and added to 3.8 ml of buffer (150 mM NaCl, 30 mM Hepes, and 10 mM dextrose; pH = 7.4). Four hundred-microliter aliquots of the latter suspension were placed in each of two small glass tubes. Fifty microliters of 500 nM ouabain in buffer was added to one tube while the same volume of 500 PM ouabain in buffer was added to the other tube. Both of these ouabain solutions also contained 1.0 pmole of 13H]ouabain (17 Ci/mmole, Amersham). The tubes were incubated at 37°C for 60 min with gentle agitation. The cells were then washed three times with ice-cold choline chloride and their bound radioactivity eluted with 300 ~1 of 12% trichloroacetic acid. The difference in counts per minute by red cells exposed to the lower (55 nM) versus the higher (55 FM) concentration of ouabain was taken as a measure of ouabain bound
234
WEBSTER,
VAN
WINKLE.
AND
KARRAT
specifically to Na’-K ‘-ATPase. A counting efficiency of 42% was used to calculate the moles of ouabain bound to the Na’-K’-ATPase contained in 20 ~1 of erythrocytes. Erythrocyte intracellular Na’ concentrations were determined by lysing 100 ~1 of the “packed” red cells prepared above in 9.9 ml of 3 PM lithium nitrate solution, then analyzing the preparation by flame atomic absorption spectroscopy. Blood pressures were measured using a conventional mercury column sphygmomanometer. Correlation coefficients were calculated to detect statistically significant correlations among ouabain binding, intracellular Na’ , and diastolic blood pressure. One-tailed tests were considered appropriate since the expected relationships were formulated before the study was begun. Changes in ouabain binding, intracellular Na’ , and diastolic blood pressure after dieting were analyzed statistically using two-tailed pairing design t tests. We used the x’ goodness of fit test to determine if data were distributed normally (8). RESULTS
The ouabain binding level was inversely correlated with diastolic blood pressure (Fig. 1: r = -0.59) and intracellular Na’ concentration (Fig. 2; r = -0.68) in patients receiving medication for their hypertension. Intracellular Nat correlated directly (r = +0.68) with diastolic blood pressure in the latter patients (Fig. 3). In contrast, there was only a weak inverse correlation (r = - 0.26) between ouabain binding and intracellular Na+ (Fig. 2), and no correlation between diastolic blood pressure and intracellular Na+ or ouabain binding among nonmedicated, normotensive obese patients (Figs. 1 and 3). This lack of relationship among the parameters was not altered when data from the 15 heaviest, 16 oldest, or 11 “heaviest and oldest” normotensive patients were considered in separate analyses. Consistent with the preceding observations, medicated patients had lower diastolic blood pressure (P < 0.05) and intracellular Na’ (P < 0.05) and higher ouabain binding (P < 0.02) after dieting. These parameters were not affected by dieting in normotensive patients (Table I). However, two populations of ouabain binding were found among the latter group (Fig. 4). DISCUSSION Blood pressure and Naf transport in normotensive patients. If erythrocyte ouabain binding reflects a level of ouabain binding characteristic of an individual, then there may be no relationship between blood pressure and smooth muscle cell Na’-K+-ATPase levels among normotensive, obese patients (Fig. 1). If Na’-K’-ATPase levels are to contribute to blood pressure via the mechanism proposed by Blaustein (4), they must
Na’ . OBESITY,
235
AND HYPERTENSION
0
60
n 10
20
n.
30 Ouobain
40 binding
(tmO~/2Oul
50
60
cells1
FIG. 1. Relationship between ouabain binding and diastolic blood pressure in obese women. Women receiving medication for hypertension (0; r = -0.59; P < 0.02; onetailed test); normotensive patients (0; r = +0.12). One (0) point (48 mm Hg) is not shown. The fine is for medicated patients.
Sect the intracellular Na+ concentration. However, only a weak inverse correlation was observed between ouabain binding and intracellular Na’ concentration (Fig. 2), suggesting that other Na+ transport mechanisms may partially compensate for the potential effect of Na’-K’-ATPase levels on intracellular Na+. Moreover, since the intracellular Na+ con-
6 90
I 40
Al 3a Ouabain
biding
n
(fmd
I 50
. w
/20uicelb)
FIG. 2. Relationship between ouabain binding and intracellular Na’ in obese women. Women receiving medication for hypertension (A; r = -0.68; P < 0.01); normotensive patients (0; r = -0.26; P < 0.05; one-tailed test). The line is for medicated patients.
236
WEBSTER,
VAN WINKLE,
lnlracellular
AND KARRAT
No+ (mmol / I cells)
FIG. 3. Relationship between intracellular Na- and diastolic blood pressure in obese women. Women receiving medication for hypertension (+; r = +0.68; P < 0.01); normotensive patients (0; r = -0.01). One (0) point (48 mm Hg) is not shown. The line is for medicated patients.
centration is unrelated to diastolic blood pressure among normotensive patients (Fig. 3), other factors may compensate for the residual effect of Na’-K+-ATPase levels on intracellular Na+, thus completely compensating for the potential effect of the enzyme levels on blood pressure (Fig. 1). Blood pressure and Na’ transport in hypertesive patients receiving treatment. A strong inverse correlation between ouabain binding levels
Ouobain
Binding
(fmol / 20~1 cells)
FIG. 4. Distribution of erythrocyte ouabain binding among obese women. The levels of ouabain binding were not normally distributed among normotensive women (open bars; P < 0.05). Cross hatching indicates patients receiving hypertensive medication.
Na’,
OBESITY,
AND
HYPERTENSION
237
and intracellular Na+ concentration was seen in obese patients receiving medication for hypertension (Fig. 2). Intracellular Na+ levels were strongly correlated with blood pressure (Fig. 3) and, as expected from these two relationships, ouabain binding was inversely correlated with blood pressure (Fig. 1). Thus it appears that obese patients receiving medication for hypertension may be less able to compensate (a) for an effect of Na+K+-ATPase levels on intracellular Na’ and (b) for an effect of intracellular Na’ levels on blood pressure. Consistent with this hypothesis, an increase in ouabain binding was accompanied by a decrease in both intracellular Na’ and blood pressure when medicated patients dieted (Table 1). One might theorize that dieting reduces Na’ intake and that lower Na’ consumption decreases production of an inhibitor of Na’-K+-ATPase (9); thus affecting ouabain binding levels in some way (other explanations for this increase in ouabain binding are also possible, as discussed below). Regardless of the reason(s) for the changes, however, an increase in ouabain binding was associated with a decrease in intracellular Na+ and blood pressure in each of the medicated patients. In contrast, only half of the normotensive patients from whom blood samples were available after dieting had changes in blood pressure and intracellular Na’ in the opposite direction of the change in their level of erythrocyte ouabain binding. This occurred in spite of the fact that the changes in ouabain binding and blood pressure after dieting were of similar absolute magnitude in both medicated and nonmedicated, normotensive patients. Thus, Na’K’-ATPase levels appear to be inversely correlated with intracellular Na’ and blood pressure only in those obese patients also receiving medication for hypertension; possibly because their compensatory mechanisms have broken down. “Primary obesity” as a possible risk factor for hypertension. Our data (Fig. 4) support the hypothesis of De Luise et al. (1,2) that there are at least two populations of Na’-K’-ATPase levels among obese people. One group appears to have a family history of obesity (primary obesity), while another group may be obese as a result of other conditions such as central nervous system lesions (secondary obesity) (1,2). Subjects with primary obesity appear to have lower than normal Na+-K’-ATPase levels which may predispose them not only to obesity (1,2), but to hypertension as well (6). Mir et al. (10) have suggested that variation in NaC-K’-ATPase activity may result, in part, from differences in food intake near the time when blood samples are obtained from subjects. Thus, we may have found two populations of ouabain binding levels among obese people because some patients had eaten before coming to our clinic and some had not. Mir and associates have also presented data that seem to contradict the findings of De Luise et al. (1,2) since they failed to detect diminished
INTRACELLULAR
4.8 2 0.4 4.5 2 0.3
Mean (t SE) duration of dieting (weeks)
BINDING,
TABLE
1
, .AND BLOOD PRESSURF METHYLWPA
0.2 -c 3.7 6.1 k 1.7”
Ouabain binding (fmole/20 ~1 of cells)
NA-
OBLSL
WOMEN
1s OBESF
0.3 2 1.3 -4.6 t 1.7
Intracellular Na* (mM)
-0.7 -7.0
2 5.5 ‘- 2.5’
Wows
-2.7 2 0.8” - 2.4 i 0.7”
Body mass (kg)
RECEIVING
Na-. dias:olic blood binding after dieting
3 years, respectively. 2 year<. respectively.
Diastolic blood pressure (mm Hg)
Mean change after dieting 2 SE
IN NORMOTENSIVE
” These normotensive patients had a mean initial body mass. weight, and age 2 SE of 89 ? 5 kg, 162 +_ 2 cm. and 33 -f * These hypertensive patients had a mean initial body mass, height, and age -+ SE of 100 + S kg. 164 I 2 cm. and 42 -+ P < 0.05. ‘! P c: 0.02 (pairing design f test; 8). Statistical tests were applied to compare parameters (ouabain binding. intracellular pressure. and body mass) before vs after dieting [e.g.. patients receiving methyldopa had significantly higher levels of ouabain (P < 0.02) while this parameter was not affected when nonmedicated patients dieted].
6 6
None” Methyldopa*
ON OUABAIN
Number of patients
OF DIETING
Hypertensive medication
EFFECT
Na’,
OBESITY,
AND
HYPERTENSION
239
Na’-K’-ATPase activity in the erythrocytes of a particular group of obese people, or among obese people in general (11). However, Mir et al. (IO,1 I) studied actual Na’-K+-ATPase activity (i.e., the rate of generation of inorganic phosphate from ATP) while De Luise and collaborators (1,2) studied ouabain binding. The latter is a measure of the number of Na’-K’-ATPase sites, and not an estimation of the ability of these sites to hydrolyze ATP. While it is possible that the capacity to hydrolyze ATP might be subject to modulation following a meal [it increases according to Mir et al., (IO)], it seems unlikely that the number of Na’-K’-ATPase molecules in the red cell membrane would change after eating. As expected, we found no increase in erythrocyte ouabain binding I hr after five, healthy, nonobese subjects each ate a breakfast consisting of two fried eggs; bacon, sausage, or ham: two slices of buttered toast and jelly; and eight ounces of orange juice (data not shown). These observations may explain why we (Fig. 4) and De Luise and associates (1,2) can distinguish two populations of Na+-K’-ATPase levels among obese people while Mir and collaborators ( lo,11 ) are unable to detect two similar populations. Taken together, the data suggest that primary obesity (1,2) is associated with a decrease in Na’-K’-ATPase sites per unit area of membrane. According to our premise, hypertension is more likely to occur in these patients, than in people with higher levels of Na’-Kf-ATPase, if they lose their ability to compensate for the potential effect of Na’-K’ATPase levels on intracellular Na’ and blood pressure. If obese patients receiving medication for hypertension have primary obesity (this is not a requirement of the present hypothesis), then it must be proposed that their medication has increased the number of Na’-K’-ATPase sites on the cell membranes of at least some patients. The medicated patients have ouabain binding levels which are spread across the entire range of binding found among normotensive patients (Fig. 4). In this regard, methyldopa binds tightly to erythrocyte membranes (12) and reverses the inhibition of red cell Naf-K’-ATPase by vanadate (13). Thus, methyldopa can influence Na+-K+-ATPase activity in at least one way. Moreover, if methyldopa affects Na+ -K’-ATPase levels, this drug might be even more effective when patients lose weight (e.g., milligrams of medication per kilograms of body mass would increase). Consistent with this possibility, ouabain binding increased when medicated patients dieted (Table I). Although the dose of methyldopa was reduced for one patient after only 1 week of dieting, ouabain binding was found to increase significantly after dieting whether or not this patient was considered in the statistical analysis. While the present and other (1,2) studies permit formulation of interesting hypotheses, the mechanism by which diminished ouabain binding sites
240
WEBSTER,
VAN WINKLE.
AND KARRAT
on membranes contribute to the development of obesity and/or hypertension must be elucidated before a causal relationship between these parameters can be accepted. This would include not only verification of Blaustein’s proposal (4), but a more complete description of (a) how we compensate for changes in our blood pressure and (b) the consequences when these compensatory mechanisms fail. SUMMARY People with “primary obesity” may be hypertensive because they have lost their ability to compensate for the effect of low Na’-K’-ATPase levels on blood pressure. In obese patients receiving hypertensive medication (n = 13), but not in normotensive nonmedicated patients (n = 42), diastolic blood pressure was inversely correlated with erythrocyte ouabain binding (P < 0.02) and directly correIated with intracellular Na’ concentration (P < 0.01). Moreover, there was a stronger inverse relationship between ouabain binding and intracellular Nat in patients receiving medication for hypertension (P < 0.01) than in normotensive patients (P < 0.05). These data suggest that patients receiving hypertensive medication may be less able to compensate than normotensive patients, (a) for the potential effect of Na’- K’-ATPase levels on intracellular Na+ and (b) for the potential effect of intracellular Na+ concentration on diastolic blood pressure. We propose that obese people with low levels of ouabain binding (primary obesity) may have an increased risk of developing hypertension if their compensatory mechanisms fail. ACKNOWLEDGMENTS We are indebted to Dr. Conrad Naleway of the American Dental Association Research Foundation for use of the atomic absorption spectrophotometer and Mr. Hwai-Nan Chou for technical assistance. We thank Dr. Gary Oltmans for valuable discussions. Thanks to Mr. Allan Campione, Dr. Wells Farnsworth, Barbara Le Breton, and Dr. Daniel Richardson for their help in preparing the manuscript. Supported by Chicago College of Osteopathic Medicine.
REFERENCES 1. De Luise, M., Blackbum, G. L., and Flier, J. S.. N. Engl. J. Med. 303, 1017 (1980). 2. De Luise, M., Rappaport, E., and Flier, J. S., Merabolism 31, 1153 (1982). 3. Garay, R. P., Dagher, G., Pemollet. M.-G,, Devynck, M.-A., and Meyer, P.. Narure (London) 284, 281 (1980). 4. Blaustein, M. P., Amer. J. Physiol. 232, Cl65 (1977). 5. Heagerty, A. M., Bing, R. F., Mimer, M., Thurston, H., and Swales, J. D., Lancer 2, 894 (1982). 6. Van Winkle, L. J., N. Eng/. .I. Med. 304, 358 (1981). 7. Guyton, A. C., ‘*Textbook of Medical Physiology.” Saunders, Philadelphia, 1981. 8. Snedecor, G. W., and Cochran, W. G., ‘*Statistical Methods.” Iowa State Univ. Press, Ames, 1%7.
Na’ , OBESITY,
AND HYPERTENSION
241
9. Marx, J. L., Science (Washington, D.C.) 212, 1255 (1981). 10. Mir, M. A., Charalambous, B. M., and Morgan, K., N. Engl. J. Med. 306, 809 (1982). 11. Mir, M. A., Charalambous, B. M., Morgan, K., and Evans, P. J., N. Engl. J. Med. 305, 1264 (1981). 12. Green, F. A., Jung, C. Y., Rampal, A., and Lorusso. D. J., C/in. Exp. Immunol. 40, 554 (1980). 13. Hudgins, P. M., and Bond, G. H., Fed. Proc. 37, 3 13 (1978).